Weekday Market Patterns

# Add this to the beginning of your load-market-data chunk

# Create a data directory if it doesn't exist
if (!dir.exists("data")) {
  dir.create("data", showWarnings = FALSE)
}

# Define constants for data processing
MAX_RETURN_VALUE <- 20  # Cap at ±20% for daily returns as a reasonable bound
MAX_REGIONAL_RETURN <- 5  # Cap at ±5% for regional averages
WINSORIZE_PERCENTILE <- 0.005  # Winsorize at 0.5% on each tail

# Create a robust cache file path system for storing data
cache_file <- file.path("data", "global_market_data_cache_5yr.rds")
cache_file_weekly <- file.path("data", paste0("global_market_data_cache_5yr_", format(Sys.Date(), "%Y_%W"), ".rds"))
cache_file_daily <- file.path("data", paste0("global_market_data_cache_5yr_", format(Sys.Date(), "%Y_%m_%d"), ".rds"))

# Function to safely read cache files
safe_read_cache <- function(file_path) {
  tryCatch({
    if (file.exists(file_path)) {
      message(paste("Attempting to load cache from:", file_path))
      data <- readRDS(file_path)
      
      # Basic validation
      if (is.data.frame(data) && nrow(data) > 0 && 
          all(c("date", "symbol", "close") %in% colnames(data))) {
        message(paste("Successfully loaded cache with", nrow(data), "rows from", file_path))
        return(data)
      } else {
        message("Cache data structure is invalid or empty")
      }
    }
    return(NULL)
  }, error = function(e) {
    message(paste("Error reading cache file:", e$message))
    return(NULL)
  })
}

# Function to safely save cache
safe_save_cache <- function(data, file_path) {
  tryCatch({
    # Create directory if needed
    dir.create(dirname(file_path), showWarnings = FALSE, recursive = TRUE)
    
    # Save data
    saveRDS(data, file_path)
    message(paste("Successfully saved cache to:", file_path))
    return(TRUE)
  }, error = function(e) {
    message(paste("Failed to save cache to:", file_path, "-", e$message))
    return(FALSE)
  })
}

market_data <- NULL

# Try to load from cache files in order of recency
if(is.null(market_data)) {
  market_data <- safe_read_cache(cache_file_daily)
}

if(is.null(market_data)) {
  market_data <- safe_read_cache(cache_file_weekly)
}

if(is.null(market_data)) {
  market_data <- safe_read_cache(cache_file)
}

# If no cache, new code for fetching data will run as defined in your existing logic

# A tibble: 10 × 2
   region          count
   <chr>           <int>
 1 Western Europe      7
 2 East Asia           5
 3 Middle East         5
 4 North America       5
 5 South Asia          5
 6 Africa              4
 7 Eastern Europe      4
 8 Latin America       4
 9 Northern Europe     3
10 Oceania             2